Considerable medical literature and clinical experience establish that ventilator associated pneumonia (VAP) is a feared and often fatal complication of mechanical ventilation. In the United States, over 250,000 patients per year are stricken with VAP or approximately 800 cases per million population. In Melbourne, the incidence in 2006 was reported as 6.2 cases per 1,000 ventilator days, similar to the rate in the United States. Sogaard O S et al., a binational cohort study of ventilator-associated pneumonia in Denmark and Australia, Scand J Infect Dis (2006), 38:256-264). The mortality of VAP averages 25%. Therefore, in patients with a poor prognosis, a VAP diagnosis is a life threatening complication.
The onset and rapid progression to VAP usually occurs after 3-5 days of mechanical ventilation and starts with initial colonization of the airway with pathogenic bacteria. The initial colonization is followed by a purulent tracheobronchitis (also known as ventilator-associated tracheobronchitis (VAT)) which rapidly progresses to VAP. VAT is considered a precursor to VAP and is characterized as tracheobronchitis without new infiltrates on the chest radiograph (Nseir, Nosocomial tracheobronchitis Current Opinion in Infectious Diseases 2009, 22:148-153). However, not all VAT progresses to VAP, and not all VAP had a VAT precursor. In addition, pneumonia in a patient on a ventilator that was acquired in the hospital and or a nursing care facility prior to intubation and start of mechanically ventilation is often from the same highly pathogenic bacteria seen in VAP. Our use of the VAP term includes these patients as they have a similar course and prognosis of patients that develop pneumonia after initiation of mechanical ventilation.
In addition to patient mortality, VAP also prolongs ICU stays and is treated with high doses of intravenous antibiotics. However, the levels of antibiotics that can be achieved in the respiratory tract with intravenous administration are severely limited and are often lower than the effective concentrations needed to treat VAP. Moreover, the continuing emergence of drug resistant organisms, particularly in hospital settings, makes treatment with intravenous antibiotics increasingly less effective. Specifically, the emergence of multidrug-resistant bacteria such as methicillin-resistant Staphylococcus aureus (MRSA), and highly virulent Gram-negative pathogens is increasing the morbidity of VAP.
Over the past twenty years, multiple investigator sponsored trials have attempted to study aerosolized antibiotics to either treat or prevent VAP. (See Palmer et al. in Critical Care Medicine 2008, 36(7):2008-2013, Wood et al. in Pharmacotherapy 2002, 22(8):972-982, and Lu et al in AJRCCM (Volume 184:106-115, 2011).) Meta-analysis of these trials shows benefit in decreasing ventilators days and improving other outcomes. Recently, Palmer et al. supra performed a randomized blinded placebo-controlled trial to determine the impact of aerosolized antibiotics on outcomes in patients with VAT and/or VAP. Forty-three patients were randomized to receive aerosolized antibiotics or placebo for 14 days. Choice of the aerosolized antibiotics was based on Gram stain of the endotracheal aspirate. Vancomycin or gentamicin were used in patients with Gram-positive and Gram-negative microorganisms, respectively. Both antibiotics were used if both Gram-positive and Gram-negative microorganisms were present. Most of the 43 patients were also treated intravenously with systemic antibiotics. The authors found aerosolized antibiotics to be associated with significantly lower rates of VAP at the end of treatment, reduced usage of systemic antibiotics, and earlier weaning of patients from the ventilator—leading to shorter stays in the ICU.
Palmer et al. also showed the advantage of a cocktail of antibiotics, specifically gentamicin and vancomycin, that have Gram negative and Gram positive respective activity in treatment of VAP and VAT as many patients are infected with both Gram negative and Gram positive bacteria. Interestingly, lower rates of antimicrobial resistance were also found in patients treated with aerosolized antibiotics, likely as suboptimal levels, commonly seen with intravenous administration, are known to promote the development of bacterial resistance.
The delivery system used by Palmer et al. was a small particle size jet nebulizer—no longer manufactured—that introduced an additional 6 L/m airflow into the airway. Such a nebulizer is incompatible with many modern ventilators because modern ventilators have sophisticated control, monitor, and feedback systems that carefully adjust airflow and pressures in the airway. A recent study by Lu, et al., compared ceftazidime and amikacin, aerosol (n=23) vs. IV (n=17) in a small Phase 2 trial in established Gram negative bacteria and VAP. After 8 days of antibiotic administration, aerosol and intravenous groups were similar in terms of successful treatment (70% vs 55%), treatment failure (15% vs 30%), and superinfection by other microorganisms (15% vs 15%). Antibiotic resistance was observed exclusively in the intravenous group. The authors concluded that aerosol antibiotics have similar efficacy to intravenous (IV) delivery and likely lead to lower rates of bacterial resistance.
The poor results of aerosol adjunctive therapy or primary antibiotic treatment in VAP is not surprising, because intravenous antibiotics penetrate poorly into the sputum. Aerosol antibiotics generally have a 100 fold higher sputum concentration than the maximum dose IV delivery and usually with one-tenth the systemic exposure. Aerosolized antibiotics are rapidly cleared from the respiratory tract and their use can provide either very high concentrations in the lungs, which is desirable to control bacteria or very low concentrations. This avoids long periods of sub-MIC antibiotic concentrations that lead to the development of resistance. However, to date, no aerosolized antibiotics for VAP or VAT have been approved by regulatory authorities.
A promising combination of Gram-negative and Gram-positive antibiotics for VAT and VAP would be the combination of an aminoglycoside and fosfomycin. (Baker U.S. Pat. No. 7,943,118 and MacCleod J Antimicrobial Chemotherapy 2009; 64:829-836). In patients with cystic fibrosis (CF) and Pseudomonas aeruginosa (a Gram negative bacteria) infections, an 80 mg fosfomycin/20 mg tobramycin dose delivered twice daily as an aerosol by a vibrating plate nebulizer (PARI® eFlow®) was effective in decreasing the bacterial burden of P. aeruginosa, and Staphylococus aureus over a 28 day treatment period (Trapnell, et al. AJRCCM 185:171-178, 2012) Other aminoglycosides may also be synergistic with fosfomycin; Cai (J of Antimicrobial Chemotherapy 64 (2009) 563-566) reported that in both an in vitro and a systemically treated rat pseudomonas infection model, fosfomycin potentiated the efficacy of amikacin to an even greater extent than tobramycin.
In spontaneously breathing patients, the importance of a well-tolerated aerosol is also known. While mild cough can be tolerated in a patient on a ventilator, coughing increases the airway pressures, putting the patient at risk for barotrauma. It is well known that hyperosmolar solutions for nebulization can cause cough. In fact, a 7% hypertonic saline solution having an osmolality of 2411 Osm/kg is used to induce cough to obtain sputum specimens or to promote airway clearance in patients spontaneously breathing with lung disease. Lower osmolality solutions still cause cough. A formulation of fosfomycin/tobramycin with an osmolality of approximately 832 osm/kg when tested in CF patients caused-noticeable coughing in 10 of 41 patients while a placebo of normal saline (Osm/kg of 310) produced coughing in only 3 of 40 patients. Wheezing, a more several measure of bronchospasm, occurred in 5 of 41 patients compared to none in the placebo group. (AMJ Respir Crit Car Med 185:171-178, 2012).
Therefore, although some combinations of antibiotics, including fosfomycin and aminoglycosides have been used, combinations for VAP and VAT have not been approved and several problems remain to be solved. First, ventilator circuits almost invariably include a humidifier to humidify the dry gas using sterile water coming from high pressure gas supplies prior to the gas entering the patient's airway. Humidification of the air leads to hygroscopic growth of the aerosol particles. Many particles grow to a large size and “rain out” in the endotracheal and ventilator tubing or, if delivered to the patient, deposit in the large airways rather than lungs. See Miller et al., Am J Respir Crit Care Med 168:1205-09 (2003). An endotracheal tube's internal diameter averages 7-8 mm, much smaller than the diameter of a typical trachea. The smaller diameter increases “rain out” of large>5 micron particles such that those aerosol particles never reach the patient. The efficiency penalty of leaving the humidification circuit on with use of jet nebulizer with an average particle size of approximately 5 microns (at the nebulizer prior to growth due to humidification) has been estimated as loss of 50% of the aerosol. (Palmer et al. in Critical Care Medicine 1998:26:31-38). To avoid this problem, the obvious resolution is to turn off the humidifier during aerosol antibiotic therapy as was done in the treatment studies noted above (Palmer, Wood, Lu, Miller, supra). This is a successful approach but carries the finite risk of a health care worker neglecting to turn humidification back on. Accordingly, hospitals, critical care facilities, and regulatory agencies would likely require specific warning devices to maintain humidification because non-humidified gas causes drying of secretions which makes recovery from pneumonia even more difficult. A method that allows continuous humidification is optimal and would increase patient safety and treatment efficacy.
Second, improved airway tolerance is needed for patients with VAT and VAP. In the recent study of Lu, et al, patients commonly breathe out of sequence from the ventilator, likely caused by irritation from the therapeutic aerosol. Eschenbacher (Eschenbacher et al., Am Rev Respir Dis 1984, 129: 211-215) published that mild asthmatic patients will cough when exposed to an aerosol without a permeant anion (such as chloride) concentration greater than 20 meq/liter even if the aerosol is isotonic. Lu utilized sterile water to reconstitute the powdered antibiotics and did not use any saline that would provide a permeant anion in his formulation. Lu's approach was to heavily sedate the patient, which is not optimal.
Third, Eschenbacher (supra) tested a hypertonic saline solution of 1232 mOsm/liter in mild asthmatic patients and showed that cough and bronchospasm were common. Pretreatment with bronchodilators could prevent bronchospasm, but could not prevent cough. As noted above, coughing in the ventilator is not desirable, as it can cause high airway pressures leading to pneumothorax, or interfere with delivery of adequate ventilation. The osmolality of a fosfomycin/tobramycin combination currently being used in clinical studies of outpatient cystic fibrosis (CF) patients is approximately 832 mOsm/liter, far above the physiologic airway osmolality of approximately 310 mOsm/liter. The high osmolality is due to the low MW of fosfomcyin, coupled with the formulation of fosfomycin as a disodium salt. The disodium salt of fosfomycin has a solubility of 50 mg/mL of water, other salts including calcium are available, but have less solubility such that concentrated formulations are not practical. The high osmolality formulation is used so that each dose can be delivered in a 2 mL solution in a vibrating mesh nebulizer so treatment time for a CF outpatient is approximately 5 minutes. More dilute solutions would require longer administration times which lead to poor compliance and potentially less efficacy. Higher osmotic concentrations have larger hygroscopic growth. Thus, if such formulations were used in a ventilator, a large amount of >5 micron particles would rain out in the tubing, and the remaining amount that was delivered would likely be irritating to the airways. A common adverse event of cough was reported in the phase 2 CF study. In fact the high dose (160 mg fosfomycin/40 mg tobramycin, same osmolality in a 4 mL solution) was very poorly tolerated in the CF study in spite of all the patients pretreated with a bronchodilator to prevent bronchospasm.
Fourth, the epidemiology of resistance and goals of therapy are very different in VAP and VAT compared to outpatient CF. Moreover, in the hospital setting, once bacterial resistance occurs, the spread of resistant bacteria between patients is rampant and epidemics are common. Furthermore, the risk of aminoglycoside-related toxicity from the cumulative long-term dose is serious because CF patients are treated chronically for years with tobramycin aerosols. See Baker et al., USP 2007/218013. In contrast, in VAT and in VAP, a patient is likely to receive only a single, two-week course of antibiotics. Therefore, when used to treat VAP or VAT, limiting the dose of an aminoglycoside in a combination product, and relying on fosfomycin to increase bacterial killing, risks the loss of efficacy of both drugs if a patient has bacteria that are fosfomycin resistant. For example, in contrast with high ratios of fosfomycin to tobramycin disclosed in Baker, et al., an optimal formulation in VAP and VAT would have enough of an aminoglycoside dose to be an independently effective antibiotic combination. However, this approach would only increase the osmolality of the formulation if the ratio of the formulation is held constant. The advantage of the fosfomycin would be to further enhance Gram negative killing, including biofilms (Cai, supra) and also to treat Gram positive bacteria, including methicillin resistant Staphylococcus aureus (MRSA). Another advantage of fosfomycin is that it reverses some of the sputum antagonism that limits the bioavailability of aminoglycosides (MacCleod, supra), (Mendelman Am Rev Respir Dis 1985, 132:761-5). Thus, even if the bacteria is fosfomycin resistant, there may be some clinical benefit to the combination by increasing the bioactive concentrations of the aminoglycoside.
One solution would be to dilute the formulations and increase the volume placed in the nebulizer for treatment. However, observation of a patient during therapy is likely to be standard protocol. ICU specialist nurses or respiratory therapists would likely be required to observe the patient during treatment adding additional costs to the therapy due to the prolonged administration time. Serious adverse events can occur during aerosol therapy, such as in Lu's study, where a patient had a cardiopulmonary arrest due to a clogged exhalation filter on the ventilator. An optimal formulation would have shorter delivery time than that of a dilute formula. Triggering the delivery during inspiration would extend treatment time but the time loss may be offset by improvement in the efficiency of delivery. Thus, a need for a treatment protocol exists wherein lower doses may be evaluated.
Accordingly, a need exists for antibiotic compositions, equipment, and treatment methods and systems to alleviate or prevent VAT and VAP despite the known challenges and the recognized risks.